Real-time damage detection and localization of damage in vehicle components using acoustic emission and machine learning

Abstract

A method for damage detection and location using acoustic emission includes receiving, via a processor, an acoustic feedback signal from a plurality of sensors embedded in a body panel of a vehicle. The processor identifies a body panel breach and a location of the body panel breach indicative of damage on the body panel. The processor identifies the body panel breach based on the acoustic feedback signal from the plurality of sensors.

Description

INTRODUCTION[0001]The subject disclosure relates to vehicle body panel damage detection, and more specifically to computer-driven detection of damage in a vehicle body panel using acoustic signatures and machine learning.[0002]Vehicles in road service often encounter environmental factors that can damage body panels or other vehicle components. For example, highway driving can kick up stones or other objects that can strike and damage a non-metallic body panel made from carbon fiber, polytetrafluoroethylene (PTFE), or another material. Other environmental factors such as road surface, vehicle panel material, and time can also weigh on the development of vehicle panel damage. Some defects may not be visible, but may still be substantial.[0003]All materials emit an acoustic emission created by subtle vibrations from environmental factors. For example, the engine vibration in a combustion engine can also vibrate body panels at various frequencies matching engine operation. Travel acros...

AI Technical Summary

Benefits of technology

[0011]In another embodiment, identifying the body panel breach and the location of the body panel breach includes identifying an acoustic signature anomaly in the acoustic feedback signal, and determining whether the acoustic signature anomaly is dispositive of the body panel breach. Responsive to determining that the acoustic signature anomaly is not dispositive of the body panel breach, the processor compares the acoustic signature anomaly to one or more stored acoustic signature anomalies during a predetermined period of time. The processor then detects a change in the acoustic signature anomaly that is indicative of the body panel breach.

[0012]In yet another embodiment, detecting the change in the acoustic signature anomaly includes determining a propagation of the body panel breach after the predetermined period of time, where the propagation of the body panel breach is indicative of an increase in a dimension of the breach with respect to time, and indicative of a location of t

Problems solved by technology

Vehicles in road service often encounter environmental factors that can damage body panels or other vehicle components.

For example, highway driving can kick up stones or other objects that can strike and damage a non-metallic body panel made from carbon fiber, polytetrafluoroethylene (PTFE), or another material.

Other environmental factors such as road surface, vehicle panel material, and time can also weigh on the de

Images